Explanatory note on the day-ahead and intraday common ...

Explanatory note on the day-ahead and intraday

common capacity calculation methodologies for the

Core CCR

4th of June 2018

EXPLANATORY NOTE CORE DA AND ID FB CCM 4TH OF JUNE 2018

of 37

CONTENT

1. Introduction ............................................................................................................................... 5

2. Flow-based capacity calculation methodology ......................................................................... 6

2.1. Inputs – see Article 21(1)(a) of the CACM Regulation .................................................................... 6 2.1.1. Methodologies for operational security limits, contingencies and allocation constraints –

see Article 23 of the CACM Regulation ............................................................................ 6 2.1.2. Flow reliability margin (FRM) ............................................................................................. 9 2.1.3. Generation Shift Key (GSK) ............................................................................................. 13 2.1.4. Remedial Action (RA) ..................................................................................................... 13

2.2. Capacity calculation approach ...................................................................................................... 14 2.2.1. Mathematical description of the capacity calculation approach ...................................... 14 2.2.2. Adjustment for minimum RAM (AMR) ............................................................................. 16 2.2.3. CNEC selection............................................................................................................... 16 2.2.4. Long term allocated capacities (LTA) inclusion ............................................................... 18 2.2.5. Rules on the adjustment of power flows on critical network elements due to remedial

actions ............................................................................................................................ 19 2.2.6. Integration of HVDC interconnectors located within the Core CCR in the Core capacity

calculation (evolved flow-based) .................................................................................... 20 2.2.7. Capacity calculation on non-Core borders (hybrid coupling) – see Article 21(1)(b)(vii) of

the CACM Regulation ..................................................................................................... 21

3. Flow-based capacity calculation process ............................................................................... 23

3.1. High Level Process flow ............................................................................................................... 23

3.2. Creation of a common grid model (CGM) – see Article 28 of the CACM Regulation .................... 24 3.2.1. Forecast of net positions ................................................................................................. 24 3.2.2. Individual Grid Model (IGM) ............................................................................................ 25 3.2.3. IGM replacement for CGM creation ................................................................................ 25 3.2.4. Common Grid models ..................................................................................................... 26

3.3. Regional calculation of cross-zonal capacity – see Article 29 of the CACM Regulation ............... 26 3.3.1. Calculation of the final flow-based domain ...................................................................... 26

3.4. Precoupling backup & default processes – see Article 21(3) of the CACM Regulation ................ 28 3.4.1. Precoupling backups and replacement process ............................................................. 28 3.4.2. Spanning ......................................................................................................................... 29 3.4.3. Precoupling default flow-based parameters .................................................................... 29

3.5. Market coupling fallback TSO input - ATC for fallback process – see Article 44 of the CACM

Regulation .................................................................................................................................... 29

3.6. Validation of cross-zonal capacity – see Article 26 and Article 30 of the CACM Regulation ........ 31

3.7. Publication of data ........................................................................................................................ 31

3.8. Monitoring and information to regulatory authorities ..................................................................... 31

4. Implementation ....................................................................................................................... 32

4.1. Timescale for implementation of the Core flow-based day-ahead capacity calculation methodology

32


of 37

GLOSSARY

AC Allocation constraint

ACER Agency for the Cooperation of Energy Regulators

AHC Advanced Hybrid Coupling

AMR Adjustment for Minimum RAM

CACM Capacity allocation and congestion management

CC Capacity calculation

CCC Coordinated capacity calculator

CCR Capacity calculation region

CE Continental Europe(an)

CGM Common grid model

CGMA Common grid model alignment

CGMAM Common grid model alignment methodology

CNE Critical network element

CNEC Critical network element and contingency

CWE Central Western Europe(an)

DA Day-ahead

D-2 Two-days ahead

D2CF Two-days ahead congestion forecast

DC Direct current

EC External constraint

EFB Evolved flow-based

EMF European merging function

ENTSO-E European network of transmission and system operators for electricity

FAV Final adjustment value

FB Flow-based

𝐹0 Expected flow without commercial exchange within the Core region

𝐹𝑒𝑥𝑝 Expected flow

𝐹𝑚𝑎𝑥 Maximum admissible power flow

𝐹𝐿𝑇𝑁 Expected flow after long term nominations

𝐹𝑟𝑒𝑎𝑙 Real flow

𝐹𝑟𝑒𝑓 Reference flow

𝐹𝑅𝑀 Flow reliability margin

𝐺𝑆𝐾 Generation shift key

HVDC High voltage direct current

ID Intraday

IGM Individual grid model

𝐼𝑚𝑎𝑥 Maximum admissible current

LT Long term

LTA Long term allocated capacities

LTN Long term nominations submitted by Market Participants based on LTA

MC Market coupling

MCP Market clearing point

MTU Market time unit

𝑁𝑃 Net position

NRA National regulatory authority


of 37

NTC Net transfer capacity

OSL Operational security limit

PNP Preliminary net position

PPD Pre-processing data

PST Phase-shifting transformer

𝑃𝑇𝐷𝐹 Power transfer distribution factor

RA Remedial action

𝑅𝐴𝑀 Remaining available margin

RAO Remedial action optimization

RES Renewable energy sources

SA Shadow auctions

SAP Single allocation platform

TS Timestamp

TSO Transmission system operator

𝑥 scalar

�⃗� vector

𝒙 matrix


of 37

1. INTRODUCTION

Sixteen TSOs follow the decision of the Agency for the Cooperation of Energy Regulators (ACER) to

combine the existing regional initiatives of former Central Eastern Europe and Central Western Europe to

the enlarged European Core region (Decision 06/2016 of November 17, 2016). The countries within the

Core CCR are located in the heart of Europe which is why the Core CCR Project has a substantial

importance for the further European market integration.

In accordance with Article 20ff. of the CACM Regulation, the Core TSOs are working on the

implementation of the day-ahead and intraday common capacity calculation methodology (hereafter the

DA CCM and ID CCM respectively). Unless otherwise stated, the description covers the day-ahead

methodology and is equally valid for the intraday methodology where indicated.

The aim of this explanatory note is to provide additional information with regard to the day-ahead and

intraday common capacity calculation methodology and relevant processes only. This paper considers

the main elements of the relevant legal framework (i.e. CACM Regulation, 714/2009, 543/2013). Chapter

2 of this document covers the day-ahead common capacity calculation methodological aspects including

the description of the inputs and the expected outputs, while Chapter 3 details the Core DA FB CC

process.


of 37

2. FLOW-BASED CAPACITY CALCULATION METHODOLOGY

2.1. Inputs – see Article 21(1)(a) of the CACM Regulation

2.1.1. Methodologies for operational security limits, contingencies and

allocation constraints – see Article 23 of the CACM Regulation

2.1.1.1. Critical network elements and contingencies

This section refers to Article 5 of the DA CCM and Article 6 of the ID CCM.

Critical network elements (CNEs) were formerly known as Critical Branches (CBs), while contingencies

were called Critical Outages (COs). The combination of a CB and a CO (formerly CBCO) is referred to as

a Critical Network Element and Contingency (CNEC).

2.1.1.2. Operation security limits


According to Article 6(1)(a)-(c) of the DA CCM, the maximum admissible current (𝐼max) is the physical

limit of a CNE determined by each TSO in line with its operational security policy. The physical limit

reflects the capability of a transmission element (such as line, circuit-breaker, current transformer or

disconnector). This 𝐼max is the same for all the CNECs referring to the same CNE. 𝐼max is defined as a

permanent or temporary physical (thermal) current limit of the CNE in kA. A temporary current limit

represents a loading that is allowed for a certain finite duration only (e.g. 115% of permanent physical

limit can be accepted during 15 minutes). Each individual TSO is responsible for deciding, in line with

their operational security policy, if a temporary limit can be used.

As the thermal limit and protection setting can vary in function of weather conditions, the 𝐼max can be

dynamic as described in Article 6(1)(a). Dynamic Line Rating can therefore be taken into account insofar

as the necessary equipment is installed; this is perceived as being the target. For lines where this not yet

the case, seasonal variations of the 𝐼max can be applied. When the equipment limiting the 𝐼max of a CNE

is not the line itself, but another installed element physically connected to the CNE (such as current

transformer, circuit breaker, disconnector) or when the line is equipped with modern high temperature

resistant conductor material, which current limit is not dependent on the ambient temperature, a constant

𝐼max needs to be applied on all market time units.

2.1.1.3. Allocation Constraints


It is the target to have the external constraint applied directly during market coupling (MC). In such a

case the global net position (exchanges over all borders and not only those in the CCR) will be limited by

the external constraint. The concept of a global net position is illustrated in Figure 1. In this example the

global net position of bidding zone A equals: 1000 (net position in bidding zone A in CCR 1) + 500

(export to CCR 2) + 500 (export to CCR 3) = 2000 MW.


of 37

Figure 1 Illustration of global net position

If it is not feasible to apply the external constraint directly in the MC, the external constraint will be

modelled as a constraint during flow-based (FB) capacity calculation. In this case, the external

constraints are easily identifiable in the published FB parameters. Indeed, their 𝑃𝑇𝐷𝐹𝑠 are

straightforward (the zone-to-slack 𝑃𝑇𝐷𝐹 for the concerned bidding area is 1 or -1 and all the other 𝑃𝑇𝐷𝐹𝑠

are set to zero, the 𝑅𝐴𝑀 being the import/export limit (after long term nominations) and can be directly

linked to the respective bidding zone. This is demonstrated in Figure 2.

Figure 2 External constraints in the FB parameters

In this example the net position of bidding zone A cannot be higher than 3500 MW (export limit), whereas

the net position of bidding zone B cannot be lower than -1500 MW (import limit).

External constraints versus 𝑭𝑹𝑴

By construction, the flow reliability margin ( 𝐹𝑅𝑀) does not allow to hedge against the situations

mentioned in Article 8 of the DA CCM and Article 9 of the ID CCM, since they only represent the

uncertainty in forecasted flow of the FB model.

Therefore, 𝐹𝑅𝑀 on the one hand (statistical approach, looking “backward”, and “inside” the FB model)

and external constraints on the other hand (deterministic approach, looking “forward”, and beyond the

limitations of the FB model) are complementary and cannot be a substitute to each other.

CCR 1

CCR 2

CCR 3500 MW

500 MW

NPA(CCR1) = 1000 MW

NPB(CCR1) = -200 MW

NPC(CCR1) = -500 MW

NPD(CCR1) = -300 MW

A B … Z

1 0 … 0

-1 0 … 0

… … … …

Original PTDF matrix

NP

A

B

⁞

Z

≤

RAM

3500

1500

…

Original RAM vector


of 37

Legal interpretation: eligible grounds for applying allocation constraints

Under CACM, allocation constraints are understood as constraints needed to keep the transmission

system within operational security limits, which are in turn defined as acceptable operating boundaries

for secure grid operation. The definition of the latter (Article 2(7) of CACM Regulation) lists inter alia

frequency limits, thermal limits and voltage limits as some of the boundaries that can be taken into

account.

CACM does not enumerate allocation constraints (ACs) in a form of a list that would allow for checking

whether a specific constraint is allowed by the Regulation. Thus, the application of provision on allocation

constraints requires further interpretation.

CACM was issued based on Regulation 714/2009 and complements that Regulation. The general

principle in Regulation 714/2009 (Art. 16.3) is that TSOs make available the maximum capacity allowed

under secure network operation standards. Operational security is explained in a footnote to annex I as

keeping the transmission system within agreed security limits. CACM rules on AC and operational

security limits (‘OSLs’) seem to regulate the same matter as Article 16.3 in greater detail. The definition

of ACs relates to OSLs, so to define what is an allocation constraint, we first need a clear idea of OSLs.

Similarly to the ‘open’ notion of allocation constraints in the CACM, the definition of OSLs (the acceptable

operating boundaries for secure grid operation such as thermal limits, voltage limits, short-circuit current

limits, frequency and dynamic stability limits) does not include an enumerative catalogue (a closed set),

but an open set of system operation characteristics defined as to their purpose – ensuring secure grid

operation. The list is indicative (using the words ‘such as’). The open-set character of the definition is

also indicated by systemic interpretation, i.e. by the usage of the term in other network codes and

guidelines.

In the Core TSOs’ point of view, systemic interpretation allows for consistent implementation of all

network codes. In this specific case, understanding operational security limits under CACM can be

complemented by applying SO GL provisions. These, in turn, require the TSOs to apply specific

limitations to ensure that generation and load schedules resulting from cross-zonal trade do not

endanger secure system operation. In sum, operational security limits cover a broad set of system

characteristics to be respected when defining the domain for cross-zonal trade. With regard to generation

and load, this is done by applying external constraints in form of import/export limits, constituting a sub-

type of allocation constraints.

How import and export limits contribute to meeting the CACM objectives

Recital 2 of CACM Regulation preamble draws a reciprocal relationship between security of supply and

functioning markets. Thanks to grid interconnections and cross-zonal exchange, member states do not

have to fully rely on their own assets in order to ensure security of supply. At the same time, however,

the internal market cannot function properly if grid security is compromised, as market trade would

constantly be interrupted by system failures, and as a result potential social welfare gains would be lost.

Recital 18 can be seen as a follow-up, drawing boundaries to ensure a Union-wide price coupling

process, namely to respect transmission capacity and allocation constraints.

For the above reasons, one of the aims of the CACM Regulation, as expressed in Article 3, is to ensure

operational security. This aim should be fulfilled insofar it does not prejudice other aims. As explained in


of 37

this methodology, allocation constraints applied by Core TSOs do not undermine other aims of CACM

Regulation.

In line with Article 23 of CACM Regulation, allocation constraints are only used to maintain the system

within operational security limits. As the transmission system parameters used for expressing operational

security limits depend on production and consumption in a given system, these specific limitations can be

related to generation and load. As such, several assumptions that are required to assess the allocation

constraints are based on local (bidding-zone) specific parameters. Since such specific limitations cannot

be efficiently transformed into maximum active power flows on individual CNEs, these are expressed as

maximum import and export constraints of bidding zones. The inability to efficiently transform these

constraints into maximum flows on CNEs is explained in the Appendix 1 of the methodology with regard

to specific external constraints.

The data on the application of external constraints will be provided to the Core NRAs in accordance with

Article 24(3)(f) of the methodology. The final list of monitoring items, possibly containing parameters

used to determine the external constraints, will be defined before the go-live in accordance with Article

24(4).

2.1.2. Flow reliability margin (𝑭𝑹𝑴)


The methodology for the capacity calculation is based on forecast models of the transmission system.

The inputs are created two days before the delivery date of electricity with available knowledge.

Therefore, the outcomes are subject to inaccuracies and uncertainties. The aim of the reliability margin is

to cover a level of risk induced by these forecast errors.

This section describes the methodology of determining the level of reliability margin per CNEC – also

called the 𝐹𝑅𝑀 – which is based on the assessment of the uncertainties involved in the FB Capacity

Calculation (CC) process. In other words, the 𝐹𝑅𝑀 has to be calculated such that it prevents, with a

predefined level of residual risk, that the execution of the MC result (i.e. respective changes of the Core

net positions) leads to electrical currents exceeding the thermal rating of network elements in real-time

operation in the CCR due to inaccuracies of the FB CC process.

The 𝐹𝑅𝑀 determination is performed by comparing the power flows on each CNEC of the Core CCR, as

expected with the FB model used for the DA MC, with the real-time flows observed on the same CNEC.

All differences for a defined time period are statistically assessed and a probability distribution is

obtained. Finally, a risk level is applied yielding the 𝐹𝑅𝑀 values for each CNEC. The 𝐹𝑅𝑀 values are

constant for a given time period, which is defined by the frequency of 𝐹𝑅𝑀 determination process in line

with the annual review requirement. The concept is depicted in Figure 3.


of 37

Figure 3: Process flow of the 𝐹𝑅𝑀 determination

For all the hours within the one-year observatory period of the 𝐹𝑅𝑀 determination, the D-2 Common Grid

Model (CGM) is modified to take into account the real-time situation of some remedial actions that are

controlled by the TSOs (e.g. PSTs) and thus not foreseen as an uncertainty. This step is undertaken by

copying the real-time configuration of these remedial actions and applying them into the historical D-2

CGM. The power flows of the latter modified D-2 CGM are computed (𝐹𝑟𝑒𝑓) and then adjusted to realised

commercial exchanges1 inside the Core CCR with the D-2 𝑃𝑇𝐷𝐹𝑠 (see section 2.2.1). Consequently, the

same commercial exchanges in Core are taken into account when comparing the flows based on the

model created in D-2 with flows in the real-time situation. These flows are called expected flows (𝐹𝑒𝑥𝑝),

see Equation 1.

�⃗�𝑒𝑥𝑝 = �⃗�𝑟𝑒𝑓 + 𝑷𝑻𝑫𝑭 × (𝑁𝑃⃗⃗⃗⃗⃗⃗�⃗�𝑒𝑎𝑙 − 𝑁𝑃⃗⃗⃗⃗⃗⃗

�⃗�𝑒𝑓)

Equation 1

with

�⃗�𝑒𝑥𝑝 expected flow per CNEC

�⃗�𝑟𝑒𝑓 flow per CNEC in the modified D-2 CGM

𝑷𝑻𝑫𝑭 power transfer distribution factor matrix of the modified D-2 CGM

𝑁𝑃⃗⃗⃗⃗⃗⃗�⃗�𝑒𝑎𝑙 realised net position per bidding zone (based on realised exchanges)

𝑁𝑃⃗⃗⃗⃗⃗⃗�⃗�𝑒𝑓 net position per bidding zone in the D-2 CGM

1 Please note that realised commercial exchanges include the trades of all timeframes (e.g. intraday) before realtime. Exchanges naturally change the

flows in the grid from the initially forecasted flows. Hence the amount of exchanges do not lead to uncertainties itself, but the uncertainty of their

flow impact, which is modelled in the GSK, is considered in the FRM.


of 37

For the same observatory period, the realised power flows are calculated using the real-time European

grid models by means of a contingency analysis. Then for each CNEC the difference between the

realised flow (𝐹𝑟𝑒𝑎𝑙) and the expected flow (𝐹𝑒𝑥𝑝) from the FB model is calculated. Results are stored for

further statistical evaluation.

In a second step, the 90th percentile of the probability distributions of all CNECs are calculated. This

means that the Core TSOs apply a common risk level of 10% i.e. the 𝐹𝑅𝑀 values cover 90% of the

historical errors. In order to let Core TSOs have an option to reflect their attitude towards risk

acceptance, Core TSOs can then either2:

directly take the 90th percentile of the probability distributions to determine the 𝐹𝑅𝑀 of each

CNEC. This means that a CNE can have different 𝐹𝑅𝑀 values depending on the associated

contingency;

only take the 90th percentile of the probability distributions calculated on CNEs without

contingency. This means that a CNE will have the same 𝐹𝑅𝑀 for all associated contingencies.

The statistical evaluation, as described above, is conducted centrally by the CCC. The 𝐹𝑅𝑀 values will

be updated every year based upon an observatory period of one year; the 𝐹𝑅𝑀 values are then fixed

until the next update. Before the first operational calculation of the 𝐹𝑅𝑀 values, Core TSOs will

determine 𝐹𝑅𝑀 values as 10% of 𝐹𝑚𝑎𝑥 calculated under representative weather conditions, unless Core

TSOs can use the 𝐹𝑅𝑀 values already in operation in existing FB MC initatives, in which case Core

TSOs shall use those values.

In case a new CNE is added, 10% of 𝐹𝑚𝑎𝑥 is used as 𝐹𝑅𝑀. In this approach it is estimated that the 10%

common risk level corresponds to the 𝐹𝑅𝑀 value being 10% of 𝐹𝑚𝑎𝑥 calculated under representative

weather conditions, as shown in Figure 4. This approach may not necessarily reflect actual risks for all

CNECs but is considered a conservative estimation before the first calculation of the 𝐹𝑅𝑀 using above

described statistical evaluation.

2 If the same CNE is shared by two TSOs, the respective TSOs will aim to align on the same FRM value.


of 37

Figure 4: An example of the relationship between accepted risk level and the FRM value

The 𝐹𝑅𝑀 values are evaluated on a yearly basis. By doing so, no seasonal variations can be captured. It

is worth noting that computing 𝐹𝑅𝑀 values on a seasonal basis rather than a yearly basis would induce

other issues. Indeed, reducing the number of the statistical samples will reduce the validity of the model.

Besides, the smaller the sample size, the larger the impact of marginal situations will be, like atypical grid

topology / production distribution. It is foreseen that the seasonal variations potentially captured in the

calculation might be overshadowed by these undesirable impacts. A solution though to allow to capture

seasonal 𝐹𝑅𝑀 values while maintaining a large number of samples would be to compute 𝐹𝑅𝑀 values on

seasonal data over several years. However, the grid is forever evolving and changing, and the 𝐹𝑅𝑀

values would no longer have a real link to the grid situation over a long simulation period.

As a conclusion it is preferred to compute 𝐹𝑅𝑀 values over a large number of samples – hourly samples

of a year – in order to avoid introducing a high variability in the results. This allows to maintain the

performance of the capacity calculation and to capture realistic grid conditions.

After computing the 𝐹𝑅𝑀 following the above-mentioned approach, TSOs may potentially apply an

“operational adjustment” before practical implementation into their CNE and CNEC definition. The

rationale behind this is that TSOs remain critical towards the outcome of the pure theoretical approach

described above, in order to ensure the implementation of parameters that make sense operationally.

For any reason (e.g. data quality issue, perceived TSO risk level), it can occur that the “theoretical 𝐹𝑅𝑀”

is not consistent with the TSO’s experience on a specific CNE. Should this case arise, the TSO will

proceed to an adjustment. It is important to note here that this adjustment can only be a reduction of the

𝐹𝑅𝑀 value, to a value set between 5% and 20% of the 𝐹𝑚𝑎𝑥 calculated under normal weather conditions.

It is not an arbitrary re-setting of the 𝐹𝑅𝑀 but an adaptation of the initial theoretical value. The

differences between operationally adjusted and theoretical values shall be systematically monitored and

justified, which will be formalized in an annual report towards Core NRAs. Eventually, the operational

𝐹𝑅𝑀 value is determined and updated once for all TSOs and then becomes a fixed parameter in the

CNE and CNEC definition until the next 𝐹𝑅𝑀 determination.


of 37

TSOs will publish a study on the development of the 𝐹𝑅𝑀 based on the first two annual 𝐹𝑅𝑀

assessments. The internal and external parallel run data cannot be used for this purpose, as this is

simulated data only, including a simulated market outcome. Or in other words: the basis of the 𝐹𝑅𝑀

determination, being a comparison between a realised and forecasted flow is not possible before the FB

capacity calculation and allocation is live.

This implies that for some Core TSOs, only two years after go-live two annual 𝐹𝑅𝑀 datasets are

available to perform this study, and only 2.5 years after go-live the study can be published. While the

study aims to support the TSOs in their effort to reduce the 𝐹𝑅𝑀 values over time, new grid utilities and

the increasing share of renewables in the electricity production will impact the 𝐹𝑅𝑀 values, potentially

leading to their increase.

2.1.3. Generation Shift Key (𝑮𝑺𝑲)


The generation shift key (𝐺𝑆𝐾) defines how a change in net position is mapped to the generating units in

a bidding zone. Therefore, it contains the relation between the change in net position of the bidding zone

and the change in output of every generating unit inside the same bidding zone.

Due to the convexity pre-requisite of the FB domain, as required by the price coupling algorithm, the 𝐺𝑆𝐾

must be constant per market time unit (MTU).

Since the generation pattern (locations) is unique for each TSO and the range of the 𝑁𝑃 shift is also

different, there is no unique formula for all Core TSOs for the creation of the 𝐺𝑆𝐾. This is elaborated

upon in Appendix 1.

The 𝐺𝑆𝐾 values are unitless. For instance, a value of 0.05 for one generating unit means that 5% of the

change of the net position of the bidding zone will be realised by this unit. Technically, the 𝐺𝑆𝐾 values

are allocated to units in the CGM. In case where a generation unit contained in the 𝐺𝑆𝐾 is not directly

connected to a node of the CGM (e.g. because it is connected to a voltage level not contained in the

CGM), its share of the 𝐺𝑆𝐾 can be allocated to one or more nodes in the CGM in order to appropriately

model its technical impact on the transmission system.

The TSOs’ aim is to apply a 𝐺𝑆𝐾 that resembles the dispatch and the corresponding flow pattern,

thereby contributing to minimizing the 𝐹𝑅𝑀.

2.1.4. Remedial Action (RA)


In principle, all remedial actions can be preventive (applied before an outage occurs and hence effective

for all CNECs) or curative, i.e. for defined CNECs only.

Some remedial actions used in the Core CCR can have an impact on other CCRs. This is applicable to

TSOs which are part of several different CCRs. If this case shall arise, the TSO defining the RA is


of 37

responsible for coordinating its use between the different CCRs for which it is provided. This coordination

will result in an efficient use of remedial actions across CCRs.

2.2. Capacity calculation approach

2.2.1. Mathematical description of the capacity calculation approach

The FB computation is a centralized calculation which delivers two main classes of parameters needed

for the definition of the FB domain: the power transfer distribution factors (𝑃𝑇𝐷𝐹𝑠) and the remaining

available margins ( 𝑅𝐴𝑀𝑠 ). The following chapters will describe the calculation of each of these

parameters.

2.2.1.1. Power transfer distribution factor (𝑷𝑻𝑫𝑭)

This section refers to Article 12(1-7) of the DA CCM and Article 13 of the ID CCM.

The 𝑃𝑇𝐷𝐹𝑠 characterize the linearization of the model. In the subsequent process steps, every change in

net positions is translated into changes of the flows on the CNEs or CNECs with linear combinations of

𝑃𝑇𝐷𝐹𝑠. The net position (𝑁𝑃) is positive in export situations and negative in import situations. The Core

𝑁𝑃 of a bidding zone is the net position of this bidding zone with regards to the Core bidding zones.

A set of 𝑃𝑇𝐷𝐹𝑠 is associated to every CNEC after each FB parameter calculation, and gives the

influence of the variations of any bidding zone net position on the CNEC. If the 𝑃𝑇𝐷𝐹 = 0.1, this means

the concerned bidding zone has 10% influence on the CNEC. Or in other words, one MW of change in

net position leads to 0.1 MW change in flow on the CNEC. The change of flow is determined by

increasing the net position of the bidding zone and reducing the net position of the slack by the same

value.

From the calculated zone-to-slack 𝑃𝑇𝐷𝐹𝑠 (single value per bidding zone), a zone-to-zone 𝑃𝑇𝐷𝐹 can be

calculated. For example, by subtracting the zone-to-slack 𝑃𝑇𝐷𝐹 of zone 𝐵 from the one of zone 𝐴 the

impact of an exchange from zone 𝐴 to zone 𝐵 on a CNE or CNEC is determined.

In the example below, a typical zone-to-slack 𝑃𝑇𝐷𝐹 matrix is given. For each CNEC there is one zone-to-

slack 𝑃𝑇𝐷𝐹 value per bidding zone. For instance, CNEC 3 has a bidding zone(A)-to-slack 𝑃𝑇𝐷𝐹 of

14.6%. It indicates that an exchange of 1 MW from bidding zone A to the slack (which can be anywhere

in the considered grid) leads to an increased loading of 0.146 MW on CNEC 3.

Figure 5: Example zone-to-slack 𝑃𝑇𝐷𝐹𝑠

Since all commercial exchanges take place from one bidding zone to the other, only the zone-to-zone

𝑃𝑇𝐷𝐹 is a suitable indicator to determine how much a CNEC is impacted by cross-border exchanges.


of 37

Using CNEC 1 as an example, its zone(A)-to-slack 𝑃𝑇𝐷𝐹 is 4.9% and its zone(B)-to-slack is 4.8%. The

zone(A)-to-zone(B) 𝑃𝑇𝐷𝐹 is computed by 4.9% - 4.8%, yielding 0.1%. Subsequently, all zone-to-zone

𝑃𝑇𝐷𝐹𝑠 can be calculated, as shown in Figure 6.

Figure 6 : Example zone-to-zone 𝑃𝑇𝐷𝐹𝑠

The last column of Figure 6 selects the maximum zone-to-zone 𝑃𝑇𝐷𝐹 per CNEC. Investigating CNEC 1

for instance, out of three cross-border exchanges, exchange A->C holds the maximum zone-to-zone

𝑃𝑇𝐷𝐹 of 8.8%, indicating that 1 MW of A->C exchange imposes 0.088 MW on this CNEC. Please note

that one may also use Equation 6 in Article 12 in the DA CCM to directly compute the maximum zone-to-

zone 𝑃𝑇𝐷𝐹𝑠. For example, the maximum zone-to-zone 𝑃𝑇𝐷𝐹 of CNEC 1 can be computated directly by

4.9% - (-3.9%) = 8.8%, being the maximum of its zone-to-slack 𝑃𝑇𝐷𝐹 minus its minimum zone-to-slack

𝑃𝑇𝐷𝐹. Comparing to the classical full computation of the zone-to-zone 𝑃𝑇𝐷𝐹𝑠 and subsequent selection

of the maximum value as indicated in Figure 6, Equation 6 in Article 12 offers higher computational

efficiency to compute maximum zone-to-zone 𝑃𝑇𝐷𝐹𝑠.

CNEC 1, in Figure 6, has a relatively low zone-to-zone 𝑃𝑇𝐷𝐹 factor for exchanges from bidding zone A

to bidding zone B: 0.1%. If CNEC 1 has 100 MW of 𝑅𝐴𝑀 available for the allocation to be used, all the

bids and offers submitted to the allocation mechanism are competing for the scarce capacity on CNEC 1.

When all the capacity on CNEC 1 is used (and the CNEC is congested after MC), additional exchanges

from bidding zone A to bidding zone B are blocked – like all other exchanges that lead to a further

loading and congestion of the CNEC -, although a 100 MW exchange from bidding zone A to bidding

zone B induces only a 0.1 MW flow on CNEC 1.

2.2.1.2. Reference flow (𝐹𝑟𝑒𝑓)

This section refers to Article 12(8) of the DA CCM and Article 13(8) of the ID CCM.

The reference flow is the active power flow on a CNE or a CNEC based on the CGM. In case of a CNE,

the 𝐹𝑟𝑒𝑓 is directly simulated from the CGM whereas in case of a CNEC, the 𝐹𝑟𝑒𝑓 is simulated with the

specified contingency. 𝐹𝑟𝑒𝑓 can be either a positive or a negative value depending on the direction of the

monitored CNE or CNEC (see Figure 7 – the 𝐹𝑟𝑒𝑓 value is 50 MW for CNEAB but -50 MW for the

CNEBA). Its value is expressed in MW.


of 37

Figure 7: Example of a reference flow for the CNEAB

2.2.2. Adjustment for minimum RAM (AMR)

This section refers to Article 13 of the DA CCM.

Core TSOs apply an adjustment to have a minimum 𝑅𝐴𝑀 available for commercial exchanges. The

application of the minimum 𝑅𝐴𝑀 adjustment kicks in when the 𝑅𝐴𝑀 – in a situation without any

commercial exchanges in the Core region – is lower than 20% of the CNEC’s 𝐹𝑚𝑎𝑥. This is illustrated in

the examples below.

Let’s imagine a CNEC with the following values

𝐹𝑚𝑎𝑥 = 2000 𝑀𝑊 Maximimum admissible flow

𝐹𝑅𝑀 = 200 𝑀𝑊 Flow reliability margin

𝐹0 = 500 𝑀𝑊 Flow in the situation without commercial exchanges within the Core CCR

The 𝐴𝑀𝑅 for the CNEC is determined with the following equation:

𝐴𝑀𝑅 = 𝑚𝑎𝑥(0.2𝐹𝑚𝑎𝑥 − (𝐹𝑚𝑎𝑥 − 𝐹𝑅𝑀 − 𝐹0); 0) = 𝑚𝑎𝑥(400 − (2000 − 200 − 500); 0) = 0

Indeed, the 𝑅𝐴𝑀 in the situation without commercial exchanges in the Core region amounts 1300 MW

and well exceeds the boundary value of 20% of the CNEC’s 𝐹𝑚𝑎𝑥 (being 400 MW). As such, the 𝑅𝐴𝑀 of

the CNEC is not enlarged: AMR = 0.

In case the 𝐹0 = 1500 𝑀𝑊, the 𝐴𝑀𝑅 for the CNEC equals:

𝐴𝑀𝑅 = 𝑚𝑎𝑥(0.2𝐹𝑚𝑎𝑥 − (𝐹𝑚𝑎𝑥 − 𝐹𝑅𝑀 − 𝐹0); 0) = 𝑚𝑎𝑥(400 − (2000 − 200 − 1500); 0)

= 𝑚𝑎𝑥(400 − 300; 0) = 100

In this situation, the 𝑅𝐴𝑀 in the situation without commercial exchanges in the Core region amounts 300

MW and is below the boundary value of 20% of the CNEC’s 𝐹𝑚𝑎𝑥 (being 400 MW). As such, the 𝑅𝐴𝑀 of

the CNEC is enlarged to 400 MW by applying an AMR of 100 MW.

The impact of the notion of minimum 𝑅𝐴𝑀, can only be assessed in conjunction with the CNEC selection

threshold. This is elaborated upon in section 2.2.3.

2.2.3. CNEC selection


CNEC selection for capacity calculation process

This section refers to Article 5(6)(a) of the DA CCM and Article 6(6)(a) of the ID CCM.


of 37

The cross-zonal sensitivity is the criterion for selecting the CNECs that are significantly impacted by

cross-zonal trade. Cross-zonal network elements are by definition considered to be significantly

impacted. The other CNECs shall have a maximum zone-to-zone 𝑃𝑇𝐷𝐹 that exceeds the threshold of

5%.

The last column of Figure 6Error! Reference source not found. selects the maximum zone-to-zone

𝑃𝑇𝐷𝐹 per CNEC. Investigating CNEC 1 for instance, out of three cross-border exchanges, exchange A-

>C holds the maximum zone-to-zone 𝑃𝑇𝐷𝐹 of 8.8%, indicating that 1 MW of A->C exchange imposes

0.088 MW on this CNEC. Comparing with the zone-to-slack 𝑃𝑇𝐷𝐹s of CNEC 1, it is clear that, although

the zone-to-slack 𝑃𝑇𝐷𝐹s of CNEC 1 are all below 5%, the impact of cross border exchanges is still

considered significant (being 8,8%, larger than 5% CNEC selection threshold). When considering the

maximum zone-to-zone 𝑃𝑇𝐷𝐹 of CNEC 4, it is clear that this CNEC does not meet the 5% threshold

criterion. This implies that the branch will not be considered for the allocation unless it is a tie line or is

deemed necessary by the relevant TSOs.

The impact of this CNEC selection threshold can only be assessed in conjunction with the notion of

minimum 𝑅𝐴𝑀, according to Article 13 of the DA CCM. This is clarified in the following example.

A CNEC 1 with a maximum zone-to-zone 𝑃𝑇𝐷𝐹 of 5% and a minimum 𝑅𝐴𝑀 of 200 MW (being 20% of an

𝐹𝑚𝑎𝑥 = 1000 MW), is able to allow for a commercial exchange of at least 200/0.05 = 4000 MW. The

wording “at least” refers to the exchange for which the maximum zone-to-zone 𝑃𝑇𝐷𝐹 holds, i.e. for other

exchanges even higher exchanges would be feasible.

A CNEC 2 with a maximum zone-to-zone 𝑃𝑇𝐷𝐹 of 10% and an identical minimum 𝑅𝐴𝑀 of 200 MW

(being 20% of an 𝐹𝑚𝑎𝑥 = 1000 MW), is able to allow for a commercial exchange of at least 200/0.10 =

2000 MW.

Assuming that we are referring to the same pair of bidding zones for the two CNECs, the example shows

that CNEC 2 is more restrictive for the potential exchange between those two bidding zones. Or in other

words: CNEC 1 can not be limiting for the exchange between the two bidding zones in the presence of

CNEC 2.

Generally speaking, the minimum 𝑅𝐴𝑀 ensures that CNECs with lower maximum zone-to-zone 𝑃𝑇𝐷𝐹s

are less likely to become presolved than CNECs with higher maximum zone-to-zone 𝑃𝑇𝐷𝐹s.

Increasing the maximum zone-to-zone 𝑃𝑇𝐷𝐹 threshold value would essentially imply setting the

minimum 𝑅𝐴𝑀 of those CNECs, which then fall below the threshold, to an infinite value. Allowing for a

higher minimum 𝑅𝐴𝑀, or for a higher maximum zone-to-zone threshold, is likely to lead to a higher

amount of costly remedial actions required in order to maintain operational security. At the same time,

the less-constrained capacity domain is supposed to allow for a higher socio-economic welfare. The

balance between those two numbers is hard to quantify.

As indicated above, in a zonal market model, the balance in between the two extremes, touched upon

below, is hard to find:

- No internal CNECs are used in the capacity calculation and allocation.

This extreme scenario seems to optimize a day-ahead (DA) market without taking internal

constraints into account. Physical reality is however, that those internal constraints are there and

need to be coped with by the TSOs. Neglecting internal CNECs in the DA market would lead to

an “optimized” dispatch of generation and load, that to a large extent needs to be redispatched


of 37

afterwards in order to deal with the internal congestions resulting from this “optimized” market

outcome.

- All internal CNECs are used in the capacity calculation and allocation.

This extreme scenario would take the whole grid into account in the capacity calculation and

allocation. Internal congestions, though not impacted by cross-border trade whatsoever, would

block the cross-border trade.

The CNEC selection process, as proposed in the CCR Core, tries to find a proper balance in this respect,

with the introduction of both the notion of minimum 𝑅𝐴𝑀 and maximum zone-to-zone 𝑃𝑇𝐷𝐹 threshold,

and is considered to be in line with Regulation 714, Annex I, Article 1(7).

It is the belief of the Core TSOs that the proposed CNEC selection process contributes to, and does not

in any way hamper the achievement of the objectives stated in Article 3 of the CACM Regulation.

Monitored CNEC selection for Remedial Action Optimization (RAO)

This section refers to Article 5(6)(b) of the DA CCM and Article 6(6)(b) of the ID CCM.

In order to prevent overloading of non-market relevant network elements (i.e. their maximum zone-to-

zone 𝑃𝑇𝐷𝐹 falls below the 5% threshold value) due to the RAO, these network elements can be included

as monitored CNECs in the RAO. A quantitative selection criterion for these RAO-monitored CNECs is

not necessary as explained in section 2.2.5.

2.2.4. Long term allocated capacities (LTA) inclusion


In order to guarantee that the LTA are possible to be realised on the DA market, Core TSOs ensure that

the FB domain as determined during the capacity calculation process includes the LTAs. This is exactly

the goal of the LTA inclusion process, which boils down to increasing the 𝑅𝐴𝑀 on some CNECs to

accommodate possible day-ahead transactions equivalent to the amount of LTA (denoted by 𝐿𝑇𝐴𝑚𝑎𝑟𝑔𝑖𝑛

in Equation 12 in Article 14). The LTA inclusion process reflects the TSOs’ need to assure that the

congestion income collected from the day ahead is sufficient to remunerate the holders of LTA.

The type of long term transmission rights, i.e. physical or financial transmission rights, has no impact on

the LTA inclusion algorithm. It does, however, have an impact on the FB domain for the single DA

coupling, because long-term nominations, to be considered through Long term nominations (LTN)

adjustment according to Article 18(1)(f) of the DA CCM (see section 3.3.1.2), only exist on borders with

physical long-term transmission rights.

In the current configuration of the Core region, there are 17 commercial borders, which means that there

are 217

=131,072 combinations of net positions, that could result from the utilization of LTA values

calculated under the framework of the FCA guideline, to be verified against the FB domain.

The method of creating virtual constraints, as applied in the CWE FB MC, and replacing the CNEs or

CNECs for which the 𝑅𝐴𝑀𝑖 is negative will not be applied because no algorithm is known for the creation

of virtual constraints in the scope of the Core CCR (12 bidding zones, i.e. 11-dimensional problem) with

acceptable computational time.

The LTA inclusion is performed automatically in the intermediate, pre-final, and final FB computations.


of 37

In theory, such artifacts are not to be used. In practice, however, resorting to the “LTA inclusion

algorithm” can be necessary in case the FB model does not allow TSOs to reproduce exactly all the

possible market conditions. For instance, the FB capacity domain is representative to the available cross-

border capacities of the D-2 CGM whereas LT capacities are calculated in multiple market conditions.

2.2.5. Rules on the adjustment of power flows on critical network

elements due to remedial actions


The coordinated application of RAs aims at optimizing cross-zonal capacity in the Core CCR. It is a

physical property of the power system that flows can generally only be re-routed and hence a flow

reduction on one CNEC automatically leads to an increase of flow on one or more other CNECs. The

RAO aims at managing this trade-off.

A preventive tap position on a phase-shifting transformer (PST), for example, changes the reference

flow 𝐹𝑟𝑒𝑓 and thus the 𝑅𝐴𝑀. If set to the optimal position, the PST can be used to enlarge 𝑅𝐴𝑀 of highly-

loaded or congested CNECs, while potentially decreasing 𝑅𝐴𝑀 on less-loaded CNECs. The RAO itself

consists of a coordinated optimization of cross-zonal capacity within the Core CCR by means of

enlarging the FB domain in the foreseen market direction. The foreseen market direction derives from the

CGMA Methodology (section 3.2.1) and will be the result of a transparent and coordinated process that

will be agreed upon amongst Core TSOs. With this process, Core TSOs aim to minimize the risk of a

possible market intervention in case the actual market direction differs from the foreseen one. The

enlargement of the domain will be performed by the RAO on one forecasted MCP given that currently

there is no algorithm which is able to combine the results of RAO on multiple sets of net positions.

The optimization is an automated, coordinated and reproducible process. TSOs individually determine

the RAs that are given to the RA optimization, for which the selected RAs are transparent to all TSOs.

Due to the automated and coordinated design of the optimization, it is ensured that operational security

is not endangered provided that selected RAs remain available also after D-2 capacity calculation in

subsequent operational planning processes and real time. The RA optimisation does not use a specific

order in selecting RAs.

All optimization constraints for curative remedial actions are applied individually for each contingency.

Core TSOs foresee to use the following constraints for curative remedial actions:

• At most two TSOs can be involved per contingency;

• At most eight curative remedial actions can be selected per contingency;

o Limited to three curative remedial actions for RTE, due to security policy.

o Limited to two curative remedial actions for PSE, due to security policy.

• At most one topological curative remedial action can be selected per contingency respectively for

Elia, MAVIR and PSE.

As a result there is no limit regarding the maximum number of curative remedial actions on the entire set

of CNECs.


of 37

When evaluating a curative remedial action for a certain contingency, all CNEs corresponding to this

contingency will be linked with this curative remedial action. Some of these CNEs may be assigned to a

TSO which does not use curative remedial actions due to local TSO risk policy. In these cases, the

curative remedial action will only be applied if it respects all CNEs security limits in the contingency

situation.

In order to prevent overloading of non-market relevant network elements due to RAO, these network

elements can be included as RAO-monitored CNECs. For each RAO-monitored CNEC 𝑖, the following

has to hold:

RAMafter RAO,i ≥ min {0

RAMbefore RAO,i − Threshold

In words:

• If 𝑅𝐴𝑀𝑏𝑒𝑓𝑜𝑟𝑒 𝑅𝐴𝑂 is positive and higher than the threshold, 𝑅𝐴𝑀𝑎𝑓𝑡𝑒𝑟 𝑅𝐴𝑂 cannot become negative;

• If 𝑅𝐴𝑀𝑏𝑒𝑓𝑜𝑟𝑒 𝑅𝐴𝑂 is positive but smaller than the threshold, or 𝑅𝐴𝑀𝑏𝑒𝑓𝑜𝑟𝑒 𝑅𝐴𝑂 is negative,

𝑅𝐴𝑀𝑎𝑓𝑡𝑒𝑟 𝑅𝐴𝑂 shall not decrease by more than the threshold. The threshold is set at 50MW per

CNEC.

The inclusion of monitored CNECs should be based upon a qualitative and experience-based

assessment of the impact of the RAO, on the flow of each CNEC, by the TSO. Should a CNEC be

included in this list whilst having a low sensitivity towards remedial actions, it won’t have any effect on the

optimization as a threshold is used to relax the optimization problem, and the variations in (relative)

margin will be smaller than this threshold, which will be transparent as the CNEC will not be a binding

constraint during RAO. In the opposite case, if a CNEC has a significant sensitivity towards remedial

actions, it will rightfully impact the RAO as the variations in its (relative) margin will likely exceed the

threshold.

This structural shadowing effect is taken advantage of in order to ensure that the RAO-monitoring

function does not hinder the RAO result in case the qualitative assessment led to a wrongly-included

CNEC in the RAO-monitored element list.

2.2.6. Integration of HVDC interconnectors located within the Core CCR in

the Core capacity calculation (evolved flow-based)

This section refers to Article 16 and Article 12(6) of the DA CCM and Article 15 and Article 13(6) of the ID

CCM.

The evolved flow-based (EFB) methodology describes how to consider HVDC interconnectors on a

bidding zone border within the FB Core CCR during capacity calculation and efficiently allocate cross-

zonal capacity on HVDC interconnectors. This is achieved by taking into account the impact of an

exchange over an HVDC interconnector on all CNEs directly during capacity allocation. This, in turn,

allows taking into account the FB properties and constraints of the Core region (in contrast to an NTC

approach) and at the same time ensures optimal allocation of capacity on the interconnector in terms of

market welfare.

There is a clear distinction between advanced hybrid coupling (AHC) and EFB. AHC considers the

impact of exchanges between two capacity calculation regions (as the case may be belonging to two

different synchronous areas) e.g. an ATC area and a FB area, implying that the influence of exchanges


of 37

in one CCR (ATC or FB area) is taken into account in the FB calculation of another CCR. EFB takes into

account commercial exchanges over the HVDC interconnector within a single CCR applying the FB

method of that CCR.

In order to achieve the integration of the HVDC interconnector into the FB process, two virtual hubs at

the converter stations of the HVDC are added. These hubs represent the impact of an exchange over the

HVDC interconnector on the relevant CNECs. By placing a 𝐺𝑆𝐾 value of 1 at the location of each

converter station the impact of a commercial exchange can be translated into a 𝑃𝑇𝐷𝐹 value. This action

adds two columns to the existing 𝑃𝑇𝐷𝐹 matrix, one for each virtual hub. This is illustrated in Figure 8.

Figure 8 EFB applied to an HVDC interconnector between bidding zones B and C within the CCR

The virtual hubs introduced by this process are only used for the modelisation of the impact of an

exchange and will not contain any bids during MC. As a result, the virtual hubs will have a global net

position of 0 MW, but their FB net position will reflect the exchanges over the interconnector.

The list of contingencies considered in the capacity allocation is extended to include the HVDC

interconnector. Therefore, the outage of the interconnector has to be modelled as a N-1 state and the

consideration of the outage of the HVDC interconnector creates additional CNE / Contingency

combinations for all relevant CNEs during the process of capacity calculation and allocation.

2.2.7. Capacity calculation on non-Core borders (hybrid coupling) – see

Article 21(1)(b)(vii) of the CACM Regulation


Capacity calculation on non-Core borders is out of the scope of the Core FB MC project. Core FB MC

just operates provided capacities (on Core to non-Core-borders), based on approved methodologies.

The standard hybrid coupling solution which is proposed today is in continuity with the capacity

calculation process already applied in CWE FB MC. By “standard”, we mean that the influence of

“exchanges with non-Core bidding zones” on CNECs is not taken into account explicitly during the

capacity allocation phase (no 𝑃𝑇𝐷𝐹 relating to exchanges between Core and non-Core bidding zones to

the loading of Core CNECs). However, this influence physically exists and needs to be taken into

account to make secure grid assessments, and this is done in an indirect way. To do so, Core TSOs

make assumptions on what will be the eventual non-Core exchanges, these assumptions being then

A

B

C

D

A B C D VH_B VH_C

… … … … … …

… … … … … …

… … … … … …

… … … … … …

… … … … … …

PTDF matrix

NP

A

B

C

D

VH_B

VH_C

≤

RAM

…

…

…

…

…

…

RAM vector

VH_B

VH_C


of 37

captured in the D-2 CGM used as a basis, or starting point, for FB capacity calculations. The expected

exchanges are thus captured implicitly in the 𝑅𝐴𝑀 over all CNECs. Resulting uncertainties linked to the

aforementioned assumptions are implicitly integrated within each CNECs 𝐹𝑅𝑀 . As such, these

assumptions will impact (increasing or decreasing) the available margins of Core CNECs.

After the implementation of the standard hybrid coupling in the Core region, the Core TSOs are willing to

work on a target solution, in close cooperation with the adjacent involved CCRs that fully takes into

account the influences of the adjacent CCR during the capacity allocation i.e. the so-called AHC concept.


of 37

3. FLOW-BASED CAPACITY CALCULATION PROCESS

3.1. High Level Process flow

This section refers to Article 4 of the DA & ID CCM.

For DA FB capacity calculation in the Core Region, the high-level process flow foreseen is depicted in

Figure 9. It shows the various processes performed by the entities involved.

Figure 9: High level process flow for Core FB DA CC

CCC means coordinated capacity calculator, and SAP means single allocation platform

For Intraday FB capacity calculation in the Core Region, the high-level process flow will be very similar to

the foreseen process flow for DA FB capacity calculation as depicted in Figure 9. Based on experience in

Process step Day Sub-process Non Core TSOs Core TSOs CCC SAP Merging Entities

1 D-2 D-2 merging preparation (X) X X

D-2 Perform NP forecasting (X) X X

D-2 Prepare IGMs (X) X X

2 D-2 D-2 merging (X) X X

D-2 Send individual TSOs data for merging (X) X

D-2 D-2 IGM merging X

3 D-2 Initial data preparation X X X

D-2 Prepare GSK X

D-2 Prepare initial list of CNECs X X

D-2 Prepare external constraint X

D-2 Prepare list of remedial actions available X

D-2 Provide LTA X X

D-2 Fallback input file delivery if reqiuired X X X

4 D-2 Initial Data Gathering (X) X X X

D-2 Receive initial data prepration inputs (X) X X X

D-2 Merging of received inputs X

5 D-2 Initial FB computation (X) X

D-2 Perform initial Flow Based computation X

D-2 Perform CNEC selection X

D-2 Update list of CNECs X

D-2 Initate Fallback(s) if required (X) X

6 D-2 / D-1 Remedial Action optimization X

D-2 / D-1 Perform Remedial optimization X

D-2 / D-1 Remedial actions selected for Core DA FB CC X

7 D-1 Intermediate data gathering X

D-1 Provide LTN X

8 D-1 Intermediate FB computation X

D-1 Determination adjustment for minimum RAM in list of CNECs X

D-1 Execution of rules for including previously allocated capacities X

D-1 Perform intermediate Flow Based computation X

9 D-1 Validation X X

D-1 Validate cross-zonal capacties X

D-1 Coordinate cross-zonal capacties with other CCRs X

D-1 Application of FAV X

D-1 Update list of CNECs and/or external constraint X

D-1 Early publication of data X X

10 D-1 Final FB computation X

D-1 Perform Final FB computation X

D-1 Provide cross-zonal capacties to Market coupling X

D-1 Initate Fallback(s) if required X X

11 D-1 Publication X X

D-1 Publication of data X X


of 37

the Central West Europe (CWE) region, it is expected that the total process flow for an intraday capacity

calculation will have a duration in the range of 1 to 3 hours. However, it is not clear yet how this will be

impacted due to the increase of bidding zones in the Core CCR which may have a negative impact on

the computation time.

Next to this the process can only be started once all required inputs can be gathered. As the TSOs

individual IGMs may not be available at intraday gate opening, TSOs may refrain from providing any

capacity until the intraday capacity calculation process has been finilized.

3.2. Creation of a common grid model (CGM) – see Article 28 of the CACM

Regulation

3.2.1. Forecast of net positions

Forecasting of the net positions in DA time-frame in Core CCR is based on a common process

established in ENTSO-e: the Common Grid Model Alignment (CGMA). This centrally-operated process

ensures the grid balance of the models used for the daily capacity calculation across Europe. The

process is described in the Common Grid Model Alignment Methodology (CGMAM)3, which is a part of

the Common Grid Model Methodology approved by all ENTSO-e TSOs’ NRAs in 8th May 2017.

The main concept of the CGMAM is presented in Figure 10 below:

Figure 10: Main concept of the CGMAM

The CGMAM input data are created in the pre-processing phase, which shall be based on the best

available forecast of the market behaviour and Renewable Energy Source (RES) generation.

3 The “All TSOs' Common Grid Model Alignment Methodology in accordance with Article 24(3)(c) of the Common Grid Model Methodology”, dated 29th

of November 2017, can be found on ENTSO-E website:

https://www.entsoe.eu/Documents/Network%20codes%20documents/Implementation/cacm/cgmm/Common_Grid_Model_Alignment_Methodology.pdf

https://www.entsoe.eu/Documents/Network%20codes%20documents/Implementation/cacm/cgmm/Common_Grid_Model_Alignment_Methodology.pdf


of 37

Pre-processing data (PPD) of CGMA are based on either an individually or regionally-coordinated

forecast. Basically, the coordinated approach shall yield a better indicator about the final 𝑁𝑃 than an

individual forecast. Therefore, TSOs in Core CCR agreed to prepare the PPD in a coordinated way.

The main concept of the coordinated approach intends to use statistical data as well as mathematical

relationships between forecasted 𝑁𝑃 and input variables. The data shall represent the market

characteristic and the grid conditions in the given time horizon. The coefficients of the model will be

tuned by archive data.

As result of the coordinated forecast the following values are foreseen:

𝑁𝑃 per bidding zone

DC flows per interconnector

Disclaimer: the details of the methodology valid for the Core CCR are under design and proof of concept

is still required.

3.2.2. Individual Grid Model (IGM)

All TSOs develop scenarios for each market time unit and establish the IGM. This means that Core

TSOs create hourly D-2 IGMs for each day. The scenarios contain structural data, topology, and forecast

of:

intermittent and dispatchable generation;

load;

flows on direct current lines.

The detailed structure of the model for the entire ENTSO-e area, as well as the content, is described in

the Common Grid Model Methodology (CGMM), which was approved by all ENTSO-e TSOs and

regulatory authorities on 8 May 2017. In some aspects, Core TSOs decided to make the agreement

more precise concerning IGMs. Additional details are presented in following paragraphs.

The Core TSOs will use a simplified model of HVDC. It means that the DC links are represented as load

or generation.

D-2 IGMs are based on the best available forecast of the market and RES generation. As regards the net

positions, the IGMs are compliant with the CGMA process, which is common for the entire ENTSO-e

area. More specifically, the IGMs are created based on coordinated preliminary net positions (PNP),

which reflect the aforementioned best available forecast.

3.2.3. IGM replacement for CGM creation

If a TSO cannot ensure that its D-2 IGM for a given market time unit is available by the deadline, or if the

D-2 IGM is rejected due to poor or invalid data quality and cannot be replaced with data of sufficient

quality by the deadline, the merging agent will apply all methodological & process steps for IGM

replacement as defined in the CGMM.


of 37

3.2.4. Common Grid models

The individual TSOs’ IGMs are merged to obtain a CGM according to the CGMM. The process of CGM

creation is performed by the merging agent and comprises the following services:

check the consistency of the IGMs (quality monitoring);

merge D-2 IGMs and create a CGM per market time unit;

make the resulting CGM available to all TSOs.

The merging process is standardized across Europe as described in the European merging function

(EMF) requirements.

As a part of this process the merging agent checks the quality of the data and requests, if necessary, the

triggering of backup (substitution) procedures (see section 3.4 below).

Before performing the merging process, IGMs are adjusted to match the balanced net positions and

balanced flows on DC links according to the result of CGMA. For this purpose 𝐺𝑆𝐾𝑠 are used.

The Core CGM represents the entire Continental European (RG CE) transmission system4. It means that

the CGM contains not only the Core IGMs for the respected time stamps but also all IGMs of the CE

TSOs not being directly involved in the Core FB CC process.

3.3. Regional calculation of cross-zonal capacity – see Article 29 of the CACM

Regulation

3.3.1. Calculation of the final flow-based domain


Once the optimal preventive and curative RAs have been determined by the RAO process, the RAs can

be explicitly associated to the respective Core CNECs (thus altering their 𝐹𝑟𝑒𝑓 and 𝑃𝑇𝐷𝐹 values) and the

final FB parameters are computed.

When calculating the final FB parameters, the following sequential steps are taken:

1. Determination of the adjustment for minimum 𝑅𝐴𝑀 (AMR, see section 2.2.2);

2. LTA inclusion (see section 2.2.4);

3. Determining the most constraining CNECs (see section 3.3.1.1);

4. LTN adjustment (see section 3.3.1.2).

4 Members of RG CE as follow: Austria (APG, VUEN), Belgium (ELIA), Bosnia Herzegovina (NOS BiH)), Bulgaria (ESO), Croatia (HOPS), Czech

Republic (ČEPS), Denmark (Energinet.dk), France (RTE), Germany (Amprion, TenneT DE, TransnetBW, 50Hertz), Greece (IPTO), Hungary

(MAVIR), Italy (Terna), Luxembourg (Creos Luxembourg), Montenegro (CGES), Netherlands (TenneT NL), Poland (PSE S.A.), Portugal (REN),

Romania (Transelectrica), Serbia (EMS), Slovak Republic (SEPS), Slovenia (ELES), Spain (REE), Switzerland (Swissgrid), The former Yugoslav

Republic of Macedonia (MEPSO).


of 37

3.3.1.1. Determining the most constraining CNECs (“presolve”)

This section refers to Article 17 of the ID CCM as well.

Given the CNEs, CNECs and ECs that are specified by the TSOs in the Core region, the FB parameters

indicate what commercial exchanges or 𝑁𝑃𝑠 can be facilitated under the DA MC without endangering

grid security. As such, the FB parameters act as constraints in the optimization that is performed by the

Market Coupling mechanism: the net positions of the bidding zones in the Market Coupling are optimized

so that the DA social welfare is maximized while respecting the constraints provided by the TSOs.

Although from the TSO point of view, all FB parameters are relevant and do contain information, not all

FB parameters are relevant for the Market Coupling mechanism. Indeed, only those constraints that are

most limiting the net positions need to be respected in the Market Coupling: the non-redundant

constraints (or the “presolved” domain). As a matter of fact, by respecting this “presolved” domain, the

commercial exchanges also respect all the other redundant constraints. The redundant constraints are

identified and removed by the CCC by means of the so-called “presolve” process. This “presolve” step

can be schematically illustrated in the two-dimensional example depicted in Figure 11.

Figure 11: CNEs, CNECs and ECs before and after the “presolve” step

In the two-dimensional example shown above, each straight line in the graph reflects the mathematical

representation of one constraint (CNE, CNEC or EC). A line indicates the boundary between allowed and

non-allowed net positions for a specific constraint, i.e. the net positions on one side of the line are

allowed, whereas the net positions on the other side would violate this constraint (e.g. overload of a

CNEC) and endanger grid security. The non-redundant or “presolved” CNEs, CNECs and ECs define the

FB capacity domain that is indicated by the yellow region in the two-dimensional figure (see Figure 11). It

is within this FB capacity domain that the commercial exchanges can be safely optimized by the Market

Coupling mechanism. The intersections of multiple constraints (two in the two-dimensional graph in

Figure 11) define the vertices of the FB capacity domain.

3.3.1.2. LTN adjustment

As the reference flow (𝐹𝑟𝑒𝑓) is the physical flow computed from the D-2 CGM, it reflects the loading of the

CNEs and CNECs given the commercial exchanges in the D-2 CGM. Therefore, this reference flow has

to be adjusted to take into account the effect of the LTN of the MTU instead. The 𝑃𝑇𝐷𝐹𝑠 remain identical

in this step. Consequently, the effect on the FB capacity domain is a shift in the solution space, as

depicted schematically in Figure 12.


of 37

Figure 12: Shift of the FB capacity domain to the LTN

Please note that the intersection of the axes depicted in Figure 12 is the nomination point.

3.4. Precoupling backup & default processes – see Article 21(3) of the CACM

Regulation

3.4.1. Precoupling backups and replacement process


In some circumstances, it can be impossible for TSOs to compute FB parameters according to the

process and principles. These circumstances can be linked to a technical failure in the tools, in the

communication flows, or in corrupted or missing input data. Should the case arise, and even though the

impossibility to compute “normally” FB parameters only concerns one or a couple of hours, TSOs have to

trigger a backup mode in order to deliver in all circumstances a set of parameters covering the entire

day. Indeed, market-coupling is only operating on the basis of a complete data set for the whole day (all

timestamps must be available).

The approach followed by TSOs in order to deliver the full set of FB parameters, whatever the

circumstances, is twofold:

First, TSOs can trigger “replacement strategies” in order to fill the gaps if some timestamps are

missing. Because the FB method is very sensitive to its inputs, TSOs decided to directly replace

missing FB parameters by using a so-called “spanning method”. Indeed, trying to reproduce the

full FB process on the basis of interpolated inputs would give unrealistic results. These spanning

principles are only valid if a few timestamps are missing (up to 2 consecutive hours). Spanning

the FB parameters over a too long period would lead to unrealistic results.

Second, in case of impossibility to span the missing parameters, TSOs will deploy the

computation of “default FB parameters”.


of 37

The flowchart in Figure 13 captures the general approach followed by TSOs:

Figure 13: Flowchart for the application of precoupling backup and default processes

3.4.2. Spanning

When inputs for FB parameters calculation are missing for less than three hours, it is possible to

compute spanned FB parameters with an acceptable risk level, by the so-called spanning method.

The spanning method is based on an intersection of previous and sub-sequent available FB domains,

adjusted to zero balance (to delete the impact of the reference program). For each TSO, the CNEs from

the previous and sub-sequent timestamps are gathered and only the most constraining ones of both

timestamps are taken into consideration (intersection). This is illustrated in Figure 14.

Figure 14: Forming the spanned domain for the missing timestamp 2

3.4.3. Precoupling default flow-based parameters

When it is not possible to span the missing parameters, i.e. if more than two consecutive hours are

missing, the computation of “default FB parameters” will be deployed.

3.5. Market coupling fallback TSO input - ATC for fallback process – see Article

44 of the CACM Regulation


As a result of FB CC, FB domains are determined for each MTU as an input for the FB MC process. In

case the latter fails, the FB domains will serve as the basis for the determination of the ATC values that

are input to the fallback process (ATCs for fallback process). In other words: there will not be a need for


of 37

an additional and independent stage of ATC capacity calculation. As the selection of a set of ATCs from

the FB domain leads to an infinite set of choices, an algorithm has been designed that determines the

ATC values in a systematic way. It is based on an iterative procedure starting from the LTA domain as

shown in Figure 15 below.

Figure 15: Creation of an ATC domain for the fallback process

The computation of the ATC for fallback process domain can be precisely described with the following

pseudo-code:

NbShares = number of Core internal commercial borders

While max(abs(margin(i+1) - margin(i))) > StopCriterionSAATC

For each constraint

For each non-zero entry in pPTDF_z2z Matrix

IncrMaxBilExchange = margin(i)/NbShares/pPTDF_z2z

MaxBilExchange = MaxBilExchange + IncrMaxBilExchange

End for

End for

For each ContractPath

MaxBilExchange = min(MaxBilExchanges)

End for

For each constraint

margin(i+1) = margin(i) – pPTDF_z2z * Max- BilExchange

End for

End While

SA_ATCs = Integer(MaxBilExchanges)

ATC domain for fallback

process


of 37

3.6. Validation of cross-zonal capacity – see Article 26 and Article 30 of the

CACM Regulation


One potential necessity for TSOs to apply an 𝐹𝐴𝑉 value during the validation of the cross-border

capacity – thereby decreasing the cross-zonal capacity – may be the need to cover significant reactive

power flows on certain CNECs. This is elaborated upon below.

Indeed, the Core TSOs explain in Article 6 of the DA CCM that they assume the share of the CNEC

loading by reactive power to be negligible. In such a case, the power factor cos(φ) = 1, which means that

the element is assumed to be loaded by active power only:

𝐹𝑚𝑎𝑥 = √3 ⋅ 𝐼𝑚𝑎𝑥 ⋅ 𝑈 ⋅ 𝑐𝑜𝑠(𝜑) = √3 ⋅ 𝐼𝑚𝑎𝑥 ⋅ 𝑈

This assumption is rather progressive and does not hold true by definition on AC grids. Normally, on the

high-voltage grids, this assumption should not pose an issue though. However, when a reactive power

flow takes substantial magnitudes, the power factor will drop according to the following equation:

𝑐𝑜𝑠(𝜑) =𝑃

√𝑃2 + 𝑄2

with

𝑃 Active power in MW

𝑄 Reactive power in Mvar

A too low value of the power factor may require the TSO to set a 𝐹𝐴𝑉 value in the validation stage. This

is illustrated in the example below.

By assuming a power factor cos(φ) = 1, a certain CNEC has the following 𝐹𝑚𝑎𝑥 to be applied in the

capacity calculation and allocation:

𝐹𝑚𝑎𝑥 = 1000 𝑐𝑜𝑠(𝜑) 𝑀𝑊 = 1000 𝑀𝑊

If the reactive power flow becomes substantial, and the power factor drops e.g. to a value cos(φ) = 0.8,

the same CNEC can only handle 800 MW of active power flow:

𝐹𝑚𝑎𝑥 = 1000 𝑐𝑜𝑠(𝜑) 𝑀𝑊 = 1000 × 0.8 𝑀𝑊 = 800 𝑀𝑊

When the TSO is not able to handle this difference of 200 MW in operations, he is able to set a 𝐹𝐴𝑉

value in the validation stage.

3.7. Publication of data


The Core transparency framework is based on the current operational transparency framework in CWE

DA FB MC.

3.8. Monitoring and information to regulatory authorities


The Core transparency framework is based on the current operational transparency framework in CWE

DA FB MC.


of 37

4. IMPLEMENTATION

4.1. Timescale for implementation of the Core flow-based day-ahead capacity

calculation methodology

This section refers to Article 25 of the DA Proposal.

According to Article 8(1) of the CACM Regulation, all European TSOs are obliged to participate in the

single day-ahead and in the single intraday coupling, thus in common implicit allocation sessions. Thus,

whilst taking into account Article 20(1) of the CACM Regulation, the target model for the TSOs of the

CCR Core is flow-based market coupling both in the day-ahead and intraday time frame. Core TSOs

strive to implement both the flow-based capacity calculation and market coupling in one single step on all

bidding zone borders.

Every other approach is neither designed yet in the current CCM proposal framework nor would it be

compliant to the aims of the CACM Regulation to ensure efficient, transparent and non-discriminatory

capacity allocation. Any approach where a capacity allocation using both implicit and explicit allocations

based on capacity domains derived out of a FB CC would need sequential approaches both in the

calculation (as net positions must be adjusted) and allocation (implicit and explicit allocations cannot be

run at the same time whilst using the same flow-based capacity inputs) that would either lead to a

potential discrimination of market actors participating in implicit allocations in the CCR or of the ones

participating in explicit auctions. Also from timing aspects such sequential solution is not feasible and

also not required by TSOs.


of 37

APPENDIX 1 - Methods for GSKs per bidding zone

The following section depicts in detail the method currently used by each Core TSO to design and

implement GSKs.

Austria:

APG’s method only considers market driven power plants in the GSK file which was done with statistical

analysis of the market behaviour of the power plants. This means that only pump storages and thermal

units are considered. Power plants which generate base load (river power plants) are not considered.

Only river plants with daily water storage are also taken into account in the GSK file. The list of relevant

power plants is updated regularly in order to consider maintenance or outages.

Belgium:

Elia will use in its GSK flexible and controllable production units which are available inside the Elia grid

(they can be running or not). Units unavailable due to outage or maintenance are not included.

The GSK is tuned in such a way that for high levels of import into the Belgian bidding zone all units are,

at the same time, either at 0 MW or at Pmin (including a margin for reserves) depending on whether the

units have to run or not (specifically for instance for delivery of primary or secondary reserves). For high

levels of export from the Belgian bidding zone all units are at Pmax (including a margin for reserves) at the

same time.

After producing the GSK, Elia will adjust production levels in all 24 hour D2CF to match the linearised

level of production to the exchange programs of the reference day as illustrated in Figure 16.

Figure 16: Belgian GSK.

Croatia:

HOPS will use in its GSK all flexible and controllable production units which are available inside the

HOPS’ grid (mostly hydro units). Units unavailable due to outage and maintenance are not included, but

units that aren’t currently running are included in GSK. In addition also load nodes that shall contribute to

the shift are part of the list in order to take into account the contribution of generators connected to lower

Max exportMax import 0

Pmax

Pmin


of 37

voltage levels (implicitly contained in the load figures of the nodes connected to the 220 and 400 kV

grid). All mentioned nodes are considered in shifting the net position in a proportional way.

Czech Republic:

The Czech GSK considers all production units which are available inside CEPS´s grid and were foreseen

to be in operation in target day. Units planned for the maintenance and nuclear units are not included in

the GSK file. The list of GSK is produced on hourly basis. The units inside the GSK will follow the change

of the Czech net position proportionally to the share of their production in the D-2 CGM. In other words, if

one unit represents n% of the total generation on the Czech bidding zone in the D-2 CGM, n% of the shift

of the Czech net position will be attributed to this unit.

The current approach of creation GSKs is regularly analysed and can be adapted to reflect actual

situation in CEPS´s grid.

Netherlands:

TenneT TSO B.V. will dispatch controllable generators in such a way as to avoid extensive and not

realistic under- and overloading of the units for foreseen extreme import or export scenarios.

Unavailability due to outages are considered in the GSK. Also the GSK is directly adjusted in case of

new power plants.

All GSK units (including available GSK units with no production in de D2CF file) are redispatched pro

rata on the basis of predefined maximum and minimum production levels for each active unit in order to

prevent unfeasible production levels.

The maximum production level is the contribution of the unit in a foreseen extreme maximum production

scenario. The minimum production level is the contribution of the unit in a foreseen extreme minimum

production scenario. Base-load units will have a smaller difference between their maximum and minimum

production levels than start-stop units.

TenneT TSO B.V. will continue fine-tuning their GSK within the methodology shown above.

France:

The French GSK is composed of all the flexible and controllable production units connected to RTE’s

network in the D-2 CGM.

The variation of the generation pattern inside the GSK is the following: all the units which are in operation

in the D-2 CGM will follow the change of the French net position based on the share of their productions

in the D-2 CGM. In other words, if one unit represents n% of the total generation on the French bidding

zone in the D-2 CGM, n% of the shift of the French net position will be attributed to this unit.

Germany:

The German5 TSOs provide one single GSK for the whole German bidding zone. Since the structure of

the generation differs for each German TSO, an approach has been developed, which allows the single

TSO to provide GSKs that respect the specific character of the generation in their own grid while

ultimately yielding a comprehensive single German GSK.

5 The area of Luxemburg is taken into account in the contribution from Amprion.


of 37

In a first step, each German TSO creates a TSO-specific GSK with respect to its own control area based

on its local expertise. The TSO-specific GSK denotes how a change of the net position in the forecasted

market clearing point of the respective TSO’s control area is distributed among the nodes of this area.

This means that the nodal factors of each TSO-specific GSK add up to 1. Details of the creation of the

TSO-specific GSKs are given below per TSO.

In a second step, the four TSO-specific GSK are combined into a single German GSK by assigning

relative weights to each TSO-specific GSK. These weights reflect the distribution of the total market

driven generation among German TSOs. The weights add up to 1 as well.

With this method, the knowledge and experience of each German TSO can be brought into the process

to obtain a representative GSK. As a result, the nodes in the GSK are distributed over whole Germany in

a realistic way, and the individual factors per node are relatively small.

Both the TSO-specific GSKs and the TSOs’ weights are time variant and updated on a regular basis.

Clustering of time periods (e.g. peak hours, off-peak hours, week days, weekend days) may be applied

for transparency and efficiency reasons.

Individual distribution per German TSO

50Hertz:

The GSK for the control area of 50Hertz is based on a regular statistical assessment of the behaviour of

the generation park for various market clearing points. In addition to the information on generator

availability, the interdependence with fundamental data such as date and time, season, wind infeed etc.

is taken into account. Based on these, the GSK for every MTU is created.

Amprion:

Amprion established a regular process in order to keep the GSK as close as possible to the reality. In

this process Amprion checks for example whether there are new power plants in the grid or whether

there is a block out of service. According to these monthly changes in the grid Amprion updates its GSK.

If needed Amprion adapts the GSK in meantime during the month.

In general Amprion only considers middle and peak load power plants as GSK relevant. With other words

base load power plants like nuclear and lignite power plants are excluded to be a GSK relevant node.

From this it follows that Amprion only takes the following types of power plants: hard coal, gas and hydro

power plants. In the view of Amprion only these types of power plants are taking part of changes in the

production.

TenneT Germany:

Similar to Amprion, TTG considers middle and peak load power plants as potential candidates for the

GSK. This includes the following type of production units: coal, gas, oil and hydro. Nuclear power plants

are excluded upfront.

In order to determine the TTG GSK, a statistical analysis on the behaviour of the non-nuclear power

plants in the TTG control area has been made with the target to characterize the units. Only those power

plants, which are characterized as market-driven, are put in the GSK. This list is updated regularly.

TransnetBW:


of 37

To determine relevant generation units, TransnetBW takes into account the power plant availability and

the most recent available information from the independent power producer at the time when the

individual GSK-file is created.

The GSK for every considered generation node i is determined as:

𝐺𝑆𝐾𝑖 = 𝑃𝑚𝑎𝑥,𝑖 − 𝑃𝑚𝑖𝑛,𝑖

∑ (𝑃𝑚𝑎𝑥,𝑖 − 𝑃𝑚𝑖𝑛,𝑖)𝑛𝑖=1

Where n is the number of power plants, which are considered for the generation shift within

TransnetBW’s control area.

Only those power plants which are characterized as market-driven, are used in the GSK if their

availability for the MTU is known.

The following types of generation units connected to the transmission grid can be considered in the GSK:

hard coal power plants

hydro power plants

gas power plants

Nuclear power plants are excluded.

Hungary:

MAVIR uses general GSK file listing all possible nodes to be considered in shifting the net position in a

proportional way, i.e. in the ratio of the actual generation at the respective nodes. All dispatchable units,

including actually not running ones connected to the transmission grid are represented in the list.

Furthermore, as the Hungarian power system has generally considerable import, not only big generation

units directly connected to the transmission grid are represented, but small, dispersed ones connected to

lower voltage levels as well. Therefore, all 120 kV nodes being modelled in the IGM are also listed

representing this kind of generation in a proportional way, too. Ratio of generation connected to the

transmission grid and to lower voltage levels is set to 50-50% at present.

Poland:

PSE present in GSK file all dispatchable units which are foreseen to be in operation in day of operation.

Units planned for the maintenance are not included on the list. The list is created for each hour. The units

inside the GSK will follow the change of the Polish net position proportionally to the share of their

production in the D-2 CGM. In other words, if one unit represents n% of the dispatchable generation on

the Polish bidding zone in the D-2 CGM, n% of the shift of the Polish net position will be attributed to this

unit.

Romania:

The Transelectrica GSK file contains flexible and controllable units which are available in the day of

operation. The units planned for maintenance and nuclear units are not included in the list. The fixed

participation factors of GSK are impacted by the actual generation present in the D-2 CGM.

Slovakia:

In GSK file of SEPS are given all dispatchable units which are in operation in respective day and hour

which the list is created for. The units planned for maintenance and nuclear units are not included in the

list. In addition also load nodes that shall contribute to the shift are part of the list in order to take into


of 37

account the contribution of generators connected to lower voltage levels (implicitly contained in the load

figures of the nodes connected to the 220 and 400 kV grid). All mentioned nodes to be considered in

shifting the net position in a proportional way.

Slovenia:

GSK file of ELES consists of all the generation nodes specifying those generators that are likely to

contribute to the shift. Nuclear units are not included in the list. In addition also load nodes that shall

contribute to the shift are part of the list in order to take into account the contribution of generators

connected to lower voltage levels (implicitly contained in the load figures of the nodes connected to the

220 and 400 kV grid). At the moment GSK file is designed according to the participation factors, which

are the result of statistical assessment of the behaviour of the generation units infeeds.

Explanatory note on the day-ahead and intraday common ...

Documents

Transcript of Explanatory note on the day-ahead and intraday common ...